Posicast control within feedback structure for a DCDC single ended primary

inductor converter in renewable energy applications

M. Kalantar
*
, S.M. Mousavi G.
Center of Excellence for Power System Automation and Operation, Department of Electrical Engineering, Iran University of Science and Technology, Narmak, Tehran, Iran
a r t i c l e i n f o
Article history:
Received 31 October 2009
Received in revised form 12 April 2010
Accepted 15 April 2010
Available online 26 May 2010
Keywords:
Posicast control
SEPIC converter
Average value model
Feedback
Feedforward
a b s t r a c t
In this paper, modeling, analysis, design, simulation and control of a single ended primary inductor con-
verter (SEPIC) are discussed for renewable energy applications. Because the traditional control methods
such as proportionalintegralderivative (PID) and classical half-cycle Posicast controllers based on feed-
forward are sensitive to noise and variations in natural frequency, a Posicast control with feedback struc-
ture is proposed and designed to reduce or rejection undesirable sensitivity greatly, to suppress
measurement noise and to eliminate the overshoot in the output response. The SEPIC converter is mod-
eled using average value modeling analysis. Dynamic modeling and simulation are accomplished using
MATLAB Simulink 7.2.
2010 Elsevier Ltd. All rights reserved.
1. Introduction
One of the most applications of converters such as AC/DC con-
verter, named rectier, DC/AC converter, named inverter, and DC/
DC converter, named chopper, is in renewable energy systems such
as autonomous wind-solar hybrid power generation system [1,2],
shown in Fig. 1. Among these converter, DCDC converters have
important roles for charging or discharging of batteries, adjusting
DC-link voltage between rectier and inverter and etc. Batteries
are employed to store superuous energy derived from wind blow-
ing and solar irradiance during windy or sunny days and then to
release during cloudy days or at nights. So, a DCDC converter is
used to charge or discharge the battery. Furthermore, they are also
implemented in fuel cell [3,4], photovoltaic and solar system in
order to extract the maximum power [5,6] and other applications
such as electric vehicle [7].
Various DCDC converters which can be applied in order to step
up and step down output voltage are cascaded buck and boost con-
verters, buck-boost converter, yback converter, Cuck converter
and SEPIC converter [8]. Two needed separate controllers and
switches are the most drawbacks in the cascaded buck and boost
converters. In the buck-boost and Cuck converters, output voltage
is inverted. Furthermore, a required transformer in a yback con-
verter instead of just an inductor increases the complexity of the
development. So, the best option for increasing and decreasing of
output voltage is single ended primary inductor converter named
SEPIC. A SEPIC is a DCDC converter allowing the output voltage
to change more than, less than, or equal to the input voltage with-
out inverting. Output voltage of a SEPIC is controlled using duty cy-
cle generated by control circuit and applied to the transistor. Thus,
the most advantage of a SEPIC over the other converters is a non-
inverted output voltage.
In order to control of a SEPIC converter, several controllers are
used such as classical PID and Posicast controllers. The Posicast
controller improves the steady state performance and damps reso-
nant behavior of responses. It causes that gain parameter of the
controller is easily determined and sensitivity to parametric uncer-
tainty and load change have been reduced. Furthermore, Posicast
control within a feedback system is proposed and utilized to damp
oscillations in lightly damped control systems. It is designed as a
feedback structure and dynamic compensator to deal overshoot
in the systemstep response. Also, it is very efcient, robust to mod-
eling uncertainty and is used to minimize vibration in various
types of systems and to suppress high frequency noise [9].
In this paper, Mathematical model and description of the SEPIC
converter based on average value modeling is accomplished in Sec-
tion 2. Section 3 illustrates the Posicast control design based on
feedback system. Posicast control with feedback structure is pro-
posed and designed to reduce or rejection undesirable sensitivity
greatly, to suppress measurement noise and to eliminate the over-
shoot in the step response of the SEPIC converter. Simulation re-
sults are discussed in Section 4. Finally, the conclusion is
presented in Section 5.
0306-2619/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apenergy.2010.04.012
* Corresponding author. Tel.: +98 2177240540 50x2665; fax: +98 2177240354.
E-mail address: kalantar@iust.ac.ir (M. Kalantar).
Applied Energy 87 (2010) 31103114
Contents lists available at ScienceDirect
Applied Energy
j our nal homepage: www. el sevi er. com/ l ocat e/ apenergy
2. Mathematical model of the SEPIC converter
The SEPIC converter is a kind of DCDC converter that the out-
put voltage magnitude is more or less than input voltage magni-
tude. The output voltage is adjusted using duty cycles applied to
the transistor. State space equations analysis is utilized for model-
ing and analyzing of the SEPIC converter shown in Fig. 2. In this
analysis, parasitic resistances of the inductors and capacitors are
ignored. Based on the MOSFET is on or off, two situates are happen
[8,9].
In the rst state, the MOSFET transistor is on and the diode
would not conduct. In this situation, the equivalent circuit is de-
picted in Fig. 3. Thus, state space equations are obtained as
following
L
1
0 0 0
0 L
2
0 0
0 0 C
1
0
0 0 0 C
2
_

_
_

_
..
K
d=dt
i
1
i
2
v
1
v
2
_

_
_

_
..
x

R
on
0 0 0
R
on
R
on
1 0
0 1 0 0
0 0 0 1=R
_

_
_

_
..
A
1
i
1
i
2
v
1
v
2
_

_
_

_
..
x

1 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
_

_
_

_
..
B
1
v
s
V
D
0
0
_

_
_

_
..
u
1
x
x
1
x
2
x
3
x
4
_

_
_

_

i
1
i
2
v
1
v
2
_

_
_

_
2
Kdx=dt A
1
xt B
1
ut 3
v
o
v
2
4
v
o

..
y
0 0 0 1
..
C
1
i
1
i
2
v
1
v
2
_

_
_

_
..
x
0 0
..
E
1
v
s
V
D
_ _
..
u
5
In the second state, the MOSFET transistor is off and the diode
will conduct. In this situation, the equivalent circuit is depicted
in Fig. 4. Therefore, state space equations are calculated as
following
L
1
0 0 0
0 L
2
0 0
0 0 C
1
0
0 0 0 C
2
_

_
_

_
..
K
d=dt
i
1
i
2
v
1
v
2
_

_
_

_
..
x

0 0 1 1
0 0 0 1
1 0 0 0
1 1 0 1=R
_

_
_

_
..
A
2
i
1
i
2
v
1
v
2
_

_
_

_
..
x

1 1 0 0
0 1 0 0
0 0 0 0
0 0 0 0
_

_
_

_
..
B
2
v
s
V
D
0
0
_

_
_

_
..
u
6
Kdx=dt A
2
xt B
2
ut 7
v
o
v
2
8
v
o

..
y
0 0 0 1
..
C
2
i
1
i
2
v
1
v
2
_

_
_

_
..
x
0 0
..
E
2
v
s
V
D
_ _
..
u
9
Using state space equations, small signal model for the control-to-
output transfer function are obtained as follow
+
_
s
v
1
L
2
L
1
C
2
C
D
Q
R
+
_
o
v
+
_
ref
v
e
Controller
PWM
Posicast or PID
controller
Fig. 2. SEPIC converter schematic.
+
_
s
v
1
L
2
L
1
C
2
C R
+
_
o
v
1
i
on
R
2
i
+
_
1
v
+
_
2
v
Fig. 3. SEPIC converter model when transistor is on and diode is off.
SPWM
AC/DC
DC/AC
Inverter
IG
A
C

L
o
a
d

G
Wind speed
DC/DC
Converter
& Control
DC/DC
Converter
& Control
Solar array
SEPIC
DC/DC
Convert
Battery Pack
SEPIC
SEPIC
Fig. 1. SEPIC application in a hybrid renewable energy system.
M. Kalantar, S.M. Mousavi G. / Applied Energy 87 (2010) 31103114 3111
v
o
s
ds

1
D
02

1 s
L
1
R
D
2
D
02
_ _
1 s
C
1
L
1
L
2
R
R
:
D
02
D
2
s
2 L
2
C
1
D
_ _
1
s
Q
1
x
01

s
2
x
01
_ _
1
s
Q
2
x
02

s
2
x
02
_ _ 10
x
0i
and Q
i
, cut-off frequency and quality factor respectively, are ob-
tained as follow
x
01

1

L
1
: C
1
C
2
D
2
D
02
_ _
L
2
C
1
C
2

_ 11
Q
1

R
x
01
: L
1
D
2
D
02
L
2
_ _ 12
x
02

1
L
2
C
1
D
2
k
C
2
D
02

1
L
1
C
1
kC
2

13
Q
2

R
x
02
L
1
L
2

C
1
C
2
x
2
01
x
2
02
14
where D is duty cycle and D
0
= 1 D.
3. Posicast control design based on feedback system
O.J.M. Smith proposed classical half-cycle Posicast control for
canceling the oscillatory behavior of linear lightly damped systems
and dealing the overshoot in step response in 1957 [11,12]. Not
only Posicast is applied in linear systems but also it is utilized in
a nonlinear system with slowly time varying such as a large exi-
ble-link robotic arm. Posicast based on feedforward control meth-
od relies on dynamic cancellation or model inversion. Thus,
sensitivity of classical Posicast control is very high. John Y. Hung
shows that the sensitivity problem can be reduced if Posicast com-
pensation is applied within a feedback system rather than in the
classical feedforward conguration [10].
Classical half-cycle, shown in Fig. 5, is the most typical types of
Posicast control identied by this name because its desired value is
one-half of the natural period of the plant. It is design based on
information of step response, depicted in Fig. 6, consisting of
damped response period (T
d
) and overshoot (d). Fig. 7 illustrates
the frequency response of Posicast control (1 + P(s)). Although clas-
sical Posicast control is able to cancel oscillatory response of a
lightly damped system but feedforward construction causes that
its sensitivity to modeling errors is high. This problem is solved
using feedback construction of Posicast control shown in Fig. 8.
The undamped natural frequency, x
n
, and the damping ratio, n,
are expressed by the following equations
x
n

1

L
1
C
1
C
2
D
2
D
02
_ _
L
2
C
1
C
2

_ 15
n
L
1

D
2
D
02
L
2
2R

L
1
C
1
C
2
D
2
D
02
_ _
L
2
C
1
C
2

_ 16
Also, damped natural period T
d
and step response overshoot d
are stated as follows
T
d

2p
x
n

1 n
2
_ 17
d exp
pn

1 n
2
_
_ _
18
1
2

d
T s
e

+ 1
G(s)
Posicast = 1+P(s)
r(t)
u (t)
y (t)
+
+
Lightly damped system
Fig. 5. Open-loop half-cycle Posicast control.
t
d
T 2 / T
d
1
+ 1
y
Fig. 6. Step response of a lightly damped system.
-20
-15
-10
-5
0
M
a
g
n
i
t
u
d
e

(
d
B
)
10
-2
10
-1
10
0
10
1
10
2
10
3
-90
-45
0
45
90
P
h
a
s
e

(
d
e
g
)
Bode Diagram
Frequency (rad/sec)
Fig. 7. Frequency response of Posicast, 1 + P(s), for d = 0.789 and T
d
= 25.8 ms.
) ( 1 s P + ) (s C
Lightly
Damped
System
G(s)
Posicast)
r(t)
u (t)
y (t)
+
+
+
n
_
c
G
Fig. 8. Utilized closed-loop feedback Posicast control.
+
_
s
v
1
L
2
L
1
C
2
C R
+
_
o
v
1
i
+
_
1
v
2
i
+
2
v
_
D
V
Fig. 4. SEPIC converter model when transistor is off and diode is on.
3112 M. Kalantar, S.M. Mousavi G. / Applied Energy 87 (2010) 31103114
Two steps implemented in designing feedback-based Posicast
control are the function P(s) and the controller C(s). P(s) is deter-
mined for the SEPIC converter using equations of (17) and (18)
and the controller C(s) is designed to compensate the combined
model [1 + P(s)]G(s). ZieglerNichols method is applied to design
the controller of C(s) stated as a pure integrator type compensator.
Thus, the Posicast transfer function will be as following equation:
G
c
s Cs 1 Ps
k
s
1
d
1 d
expsT
d
=2 1
_ _
19
Frequency response of the Posicast compensated function
(G
c
(s)) is shown in Fig. 9.
4. Simulation results and discussions
MATLAB Simulink 7.2 is utilized to evaluate the performance
of the Posicast controllers for the SEPIC DCDC converters. A band-
limited white noise is generated in order to consider the effects of
measurement noise. Fig. 10 shows the implemented measurement
noise. The SEPIC converter parameters are shown in Table 1.
Parameters of the Posicast control is calculated as T
d
= 25.8 ms
and d = 0.789 using Eqs. (17) and (18).
In order to get the proper response without overshoot in the
simulation, the compensator gain K is chosen as 24.5. Using Eq.
(19), the Posicast transfer function is characterized by
1 Ps 1
d
1 d
e
ST
d
=2
1 % 1 0:44e
12:9S
1 20
Output voltage for two amount of input is depicted in Fig. 11.
The rst input, V
in
= 120 V, is less than reference voltage,
V
ref
= 140 V, and the second input voltage, V
in
= 160 V, is more than
the reference voltage. In order to compare with classical controller,
a PID type feedback controller is implemented and designed. Zeros
of the PID controller are designed to cancel the lightly damped.
Output voltage is plotted in Figs. 12 and 13 for various values of
the input voltage V
in
= 130 V and V
in
= 150 V.
0 0.01 0.02 0.03 0.04 0.05
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
M
e
a
s
u
r
e
m
e
n
t

n
o
i
s
e

(
v
o
l
.
)
Time (s)
Fig. 10. Band-limited white noise signal.
Table 1
Parameters of the SEPIC converter.
Symbol Parameter Value Unit
L
1
Filter inductance 2500 lH
L
2
Filter inductance 1250 lH
C
1
Filter capacitance 2500 lF
C
2
Filter capacitance 1200 lF
R Load resistance 10 X
T
d
Damping natural period 25.8 ms
d Open loop overshoot 0.789
R
on
Transistor resistance 0.1 X
V
s
Input voltage 100180 V
f
s
Switching frequency 20 kHz
K Gain in compensator C(s) 24.5
0 0.01 0.02 0.03 0.04 0.05
0
20
40
60
80
100
120
140
160
180
Time (s.)
O
u
t
p
u
t

v
o
l
t
a
g
e

(
v
o
l
.
)
Vout for Vin=160
Vout for Vin=120
Fig. 11. Output voltage of the SEPIC for various input voltages.
-60
-40
-20
0
20
40
60
80
M
a
g
n
i
t
u
d
e

(
d
B
)
10
-2
10
-1
10
0
10
1
10
2
10
3
-180
-135
-90
-45
0
P
h
a
s
e

(
d
e
g
)
Bode Diagram
Frequency (rad/sec)
Fig. 9. Frequency response of proposed Posicast, C(s) [1 + P(s)], for d = 0.789 and
T
d
= 25.8 ms.
0 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
Time (s)
O
u
t
p
u
t

v
o
l
t
a
g
e

(
v
o
l
.
)
Posicast
PID
for Vin = 130 volt
Fig. 12. Performance of the Posicast and PID controllers on output voltage for
V
in
= 130 V.
M. Kalantar, S.M. Mousavi G. / Applied Energy 87 (2010) 31103114 3113
The response due to the proposed controller is depicted with a
solid curve (blue curve)
1
.
The response due to PID control is plotted with the dotted curve
(red curve). According to this gure, the PID controller has faster
than in rise time response in comparison with the Posicast control,
and can be tuned to give even faster response, but the PID control-
ler cannot delete the high overshoot in the output response. Fur-
thermore, responses of Posicast and PID for various values of
load resistances (R) are shown in Fig. 14. It indicates that the pro-
posed Posicast controller maintains good transient response while
load resistance changes.
Finally, the effect of measurement noise, simulated by additive
band-limited white noise, is shown in Fig. 15. The output signal
calculated by the PID controller is much noisier than the proposed
Posicast controller because the PID controller has greater high fre-
quency gain. Thus, the PID-based feedback controller does not per-
form on suppression of noise but the Posicast control based on
feedback suppress the noise properly.
5. Conclusion
In this paper, a Posicast control based on feedback structure has
been designed and implemented to eliminate the peak overshoot
output voltage of the SEPIC DCDC converter, to improve the set-
tling time of step response and to reduce the sensitivity of classical
feedforward Posicast control. T
d
and d, the Posicast elements have
been determined using the dynamics of the SEPIC converter. The
measurement noise has been highly suppressed using the pro-
posed Posicast controller while the PID controller cannot cancel
the noise effect because of a lower gain at a higher cross over fre-
quency. An integral compensator with a gain K accompanied by the
Posicast element is utilized to ensure the proper steady state re-
sponse. In Comparison with PID control, the propounded Posicast
control improved the gain and phase margins. Its narrowopen loop
bandwidth caused that the high frequency noise has been
suppressed. The proposed system maintains a suitable transient
response while load change in a wide range.
References
[1] Kalantar M, Mousavi G SM. Dynamic behavior of a stand-alone hybrid power
generation system of wind turbine, microturbine, solar array and battery
storage. Appl Energy 2010;87:305164.
[2] Gonzalez LG, Garcera G, Figueres E, Gonzlez R. Effects of the PWM carrier
signals synchronization on the DC-link current in back-to-back converters.
Appl Energy 2010;87:24919.
[3] Marsala Giuseppe, Pucci Marcello, Vitale Gianpaolo, Cirrincione Maurizio,
Miraoui Abdellatif. A prototype of a fuel cell PEM emulator based on a buck
converter. Appl Energy 2009:2192203.
[4] Hwang Jonq-Chin, Chen Li-Hsiu, Yeh Sheng-Nian. Comprehensive analysis and
design of multi-leg fuel cell boost converter. Appl Energy 2007:127488.
[5] Kang Feel-soon, Park Sung-Jun, Cho Su Eog, Kim Jang-Mok. Photovoltaic power
interface circuit incorporated with a buck-boost converter and a full-bridge
inverter. Appl Energy 2005:26683.
[6] Di Piazza Maria Carmela, Vitale Gianpaolo. Photovoltaic eld emulation
including dynamic and partial shadow conditions. Appl Energy
2010;87(3):81423.
[7] Villa Juan Luis, Sallan Jesus, Llombart Andres, Sanz Jose Fco. Design of a high
frequency Inductively Coupled Power Transfer system for electric vehicle
battery charge. Appl Energy 2009:35563.
[8] Ridley Ray. Analyzing the SEPIC converter. Power Syst Des Eur
2006:148.
[9] Erickson RW. Fundamentals of power electronics. USA: Kluwer Academic
Publishers; 1997.
[10] Feng Qi, Nelms RM, Hung JY. Posicast-based digital control of the buck
converter. IEEE Trans Ind Electron 2006;3:75967.
[11] Cook Gerald. An application of half-cycle Posicast. IEEE Trans Automat Control
1966:5569.
[12] Sugiki Akihiko, Furuta Katsuhisa. Posicast control design for parameter-
uncertain plants. In: 45th IEEE conference on decision & control, San Diego, CA,
USA; 2006. p. 31927.
0 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
Time (s)
O
u
t
p
u
t

v
o
l
t
a
g
e

(
v
o
l
.
)
R=5 Ohm
R=20 Ohm
R=30 Ohm
Fig. 14. Performance of the Posicast controller on output voltage for various values
of load resistances.
0 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
Time (s)
O
u
t
p
u
t

v
o
l
t
a
g
e

(
V
)
Posicast
PID
Fig. 15. Performance of the Posicast and PID controllers on output voltage for input
voltage coordinated with a band-limited white noise signal.
0 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
Time (s)
O
u
t
p
u
t

v
o
l
t
a
g
e

(
v
o
l
.
)
PID
Posicast
for Vin = 150
Fig. 13. Performance of the Posicast and PID controllers on output voltage for
V
in
= 150 V.
1
For interpretation of color in Figs. 1, 7, 9, 1015, the reader is referred to the web
version of this article.
3114 M. Kalantar, S.M. Mousavi G. / Applied Energy 87 (2010) 31103114